Cubic convolution for one-pass restoration and resampling

This paper describes an efficient filter that implements resampling and restoration of digital images in a single pass. The approach is to derive optimal parameters for a piecewise-cubic, reconstruction function based on a comprehensive model of the imaging process that accounts for acquisition blurring, sampling, noise, and reconstruction artifacts. For many data products produced from remotely sensed images, piecewise-cubic reconstruction is the standard algorithm for resampling. For these products, restoration with the optimal piecewise-cubic function requires virtually no additional processing resources and can significantly improve radiometric fidelity.